maryland experience in using strut-and-tie model in infrastructures

7/28/2019 Maryland Experience in Using Strut-And-tie Model in Infrastructures

1/15

Fu, Sircar and Robert 1

Paper Title: MARYLAND EXPERIENCE IN USING STRUT-AND-TIE MODEL

IN INFRASTRUCTURES

Author: Chung C. Fu, Ph.D., P.E., Director/Associate Professor, the Bridge Engineering

Software & Technology (BEST) Center

Department of Civil & Environmental Engineering

University of Maryland

College Park, MD 20742

Tel: 301-405-2011

Fax: 301-314-9129

E-mail: [email protected]

Mita Sircar, P.E.

Bechtel Corporation, Frederick, MD

Jeff Robert, P.E.

Maryland State Highway Administration, Baltimore, MD

Call for Paper:

AFF30 Concrete Bridges

Title of Session:

Application of Strut-and-Tie Model

Length:

Text 3395 words

Tables (1) 500 words

Figures (7) 3500 words

--------- ------------------------------------------

Total 7395 words


2/15


MARYLAND EXPERIENCE IN USING STRUT-AND-TIE MODEL IN

INFRASTRUCTURES

By Chung C. Fu, University of Maryland, Mita Sircar, Bechtel Corp.

and Jeff Robert, Maryland State Highway Administration

ABSTRACT

The truss model is a useful model for concrete beams failing in shear with web reinforcement.

This applies to slender beams as well as deep beams. The Strut-and-Tie Model (STM) illustrates

the powerful truss concept for reinforced concrete structures in which the compressive stresses

are resisted by the concrete struts and the tensile stresses by the reinforcing ties. Five case studies

presented here demonstrate the usage of STM in the transportation-related field. The first four

cases are simulated by planar STM models and can be solved by hand calculations or the

computer program CAST by Kuchma (2004). The fifth case describes curved concrete box girder

bridges under torsion. This case of three-dimensional torsional action can be solved by using an

extended Strut-and-Tie Model, called softened truss model, to consider both equilibrium and

compatibility. This paper focuses on demonstration of the Strut-and-Tie Model used in thetransportation area. The State of Maryland is taking this opportunity to share our STM

experience with the transportation community.


3/15


4/15


consideration of the condition of the compressed concrete at ultimate. The prerequisites of such

assumptions are:

STM is a strength design method and the serviceability should also be checked

Equilibrium must be maintained

Tension in concrete is neglected

Forces in struts and ties are uni-axial

External forces apply at nodes

Prestressing is treated as a load

Detailing for adequate anchorage is provided

In strut-and-tie truss models only equilibrium and yield criteria need to be fulfilled as the first

two requirements. But, the third requirement, the strain compatibility, is not considered. As a

result of this relaxation, more than one admissible Strut-and-Tie Model may be developed for each

load case as long as the selected truss is in equilibrium with the boundary forces and the stresses in

the Struts, Ties & Nodes are within acceptable limits.

With such a convenient structural analysis tool, questions in STM applications remain:

How to construct a Strut-and-Tie model?

If a truss can be formulated, is it adequate or is there a better one?

If there are two or more trusses for the same structure, which one is better?

Several empirical rules that provide aid in generating STM models are given below:

Elastic stress contours generated by the finite element analysis provide the generaldirection of the stress trajectories which are useful in laying out a Strut-and-Tie model.

Minimum steel content is a goal to achieve. Loads are transmitted by the principle ofminimum strain energy. Since the tensile ties are more deformable than the compression

struts, the least and shortest ties are the best.

The crack pattern may also assist in selecting the best Strut-and-Tie model. It is suggestedby the tests (MacGregor, 1997) that a STM developed with struts parallel to the orientation

of initial cracking will behave very well.

Other than the empirical rules, the common constraints are the code requirement. ACI andAASHTO code comparisons will also be discussed.

The following case studies will demonstrate usage of STMs in the transportation-related field. The

first four cases, which can be simulated by planar STM models, were solved earlier by hand

calculations and later solved by CAST (Computer-Aided Strut-and-Tie), a program developed by

Kuchma (2004). The fifth case describes curved concrete box girder bridges under torsion. This

case of three-dimensional torsional action can be solved by using an extended Strut-and-Tie Model,

called softened truss model, to consider both equilibrium and compatibility.

CASE STUDY OF BRIDGE SUBSTRUCTURE USING STM

Case Study 1 - Abutment on Pile: An abutment on piles is widely used in transportation

structures. For this case study, the abutment considered is 33-ft long, 3-ft wide and 3-ft deep.

Eleven prestressed concrete deck beams bearing on elastomeric pads are supported at intervals of

3-ft along the length of the abutment. The concrete slabs span 50-ft and transfer 107.61 kips


5/15


factored load on each elastomeric pad. The abutment is supported on 6 piles spaced at 6-ft on

center. With this geometry, where depth is just half the distance between the supports, this

abutment is a special deep beam where Bernoullis region does not exist and there is a disturbed

region throughout. AASHTO states that Bernoullis region does not exist when the depth to span

ratio exceeds 2/5. This beam just exceeds this limit. According to one of the criteria of St.

Venants principle, D-regions are those parts of a structure within a distance equal to the beam

depth of the member from the concentrated force (load or reaction).

Elevation views of the structure and Truss model are presented in Figs. 2a and 2b,respectively.Based on the calculation by CAST program, maximum compression in the diagonal strut is 101.87

kips and vertical strut is 107.61 kips. Maximum tension in the top tie is 31.76 kips and in the

bottom tie is 50.87 kips. Size of the upper nodes is determined by the size of bearing and the size

of the lower nodes is decided by the size of piles. Rebar sizes and arrangements are finalized after

a few iterations. Bearing reinforcement details in the width direction can be determined by asimple truss model in the horizontal direction. The abutment is 3-ft wide and the strut section

36x6 provides the required strength for the struts. For Ties, 3-# 6 bars can provide the required

strength. However, code-specified minimum reinforcement must be provided to prevent

temperature, creep and shrinkage related issues.

Case Study 2 - Walled Pier: Another common element found in transportation structures is a

solid shaft bridge pier on a Mat Foundation. This case study is done for an 18-ft high by 3-ft wide

wall on mat foundation. Four girders are resting on the wall and each girder reaction is 215.22 kips.

St. Venants Principle states, The localized effects caused by any load acting on the body will

dissipate or smooth out within regions that are sufficiently away from the location of the load.

Elevation of the structure is shown in Fig. 3a.

Based on the same principal, an STM model is developed for the walled pier and presented in

Fig. 3b. The inclined angle can either be obtained from stress trajectory plot or be assumed to

vary from 65 forl/d = 1 to 55 forl/d=2.0, where lis the wall length and d is the height. A

reasonable path at a 2 to 1 slope is created here to flow the concentrated loads from the top of the

wall and make their way towards the mat foundation. Maximum strut force is 128.9 kips andmaximum tie force is 50.22 kips which are in the same range of Case Study 1 and a similar strut

width and reinforcement will be sufficient. Again, for this case, minimum steel per code

provisions applicable to wall has to be provided.

Case Study 3 - Crane Beam: A conservative estimate of the resistance of a concrete structure

may be obtained by the application of the lower-bound theorem of plasticity. If sufficient ductility

is present in the system, a strut-and-tie model fulfills the conditions for the application of the above

theory. The lower-bound theorem requires identifying at least one plausible load path and insuring

that no portion of the load path is overstressed.

This case study pertains to the Gantry Crane Beam at the Maryland Port Authority Harbor (Fig.

4). The beam section is 6-ft deep by 2-ft wide and has 5-spans, each 6-ft. 135# gantry rail oncontinuous base plate (1/2-in thick by 24-in wide), anchored with the beam and the whole assembly

is encased except for the top 1-in of the rail for wheel movement. A schematic sketch of the

structure can be seen in Fig. 5.

Five-span continuous beam models are built with five different configurations to simulate the

stress trajectories for the moving wheel loads of the crane. Five configurations represent the first

wheel placed at 0, L/5, 2L/5, 3L/5, 4L/5 from the end support and other wheels follow the location

of the wheel spacing. As shown in Fig. 5, crane loads are applied at the top of the deep beam and


6/15


the self-weight of the deep beam is considered as loads to the deck. Crane load consists of 8

wheels, each 180.5 kips (factored). The envelope results for each case are tabulated in the study

report to Maryland Port Authority (Fu, 1994). The maximum tension force is 61.45 kips for Case

No. 4B and the maximum compression force is 201.84 kips, also for Case 4B. Beam thickness is

24. Based on wheel contact width and height of rail, the width of strut will be 10 minimum,

hence the strut section considered is 10x24. Reinforcements 4- #6 are provided at top and

bottom for the tie members. Truss forces and stress interaction (actual/allowable) ratios are wellbelow unity for all the members.

After achieving the solution for members, a detailed nodal analysis is performed. With 10-in

width struts the node at the bottom end of the most heavily loaded members was overstressed. A

few iterations were necessary to optimize the strut width (ranging from 10-in to 12-in) so that the

stress triangles within the nodal zone get re-oriented and meet the strength requirement of the code

specified limit of the nodal zone.

The stress fields in Struts and Ties are idealized to be uniaxial whereas the stress fields in

Nodal Zones are biaxial. These conditions cause stress discontinuity at the interface of the Strut

and Node stress fields and at the interface of the Tie and Node stress fields. The stress

discontinuity also occurs along the longitudinal boundary of Strut or Tie stress fields if the selectedstress distribution across the Effective Width is uniformly distributed. For two-dimensional

structures, the interface between two different stress fields is commonly referred to as Line of

Stress Discontinuity. Although the term "Line" is used, the stress discontinuity actually occurs on

a surface perpendicular to the plane of the structures, across the D-Region thickness. For this

reason reinforcement is required at the nodal locations perpendicular to the plane of the structures.

This reinforcement can be seen in Figure 2 provided for the case 1 example.

Case Study 4 Hammerhead Pier of Thomas Jefferson Bridge: This structure is located in

St.Marys and Calvert counties in Southern Maryland. It was completed and put into service in

1977. During an inspection in 1979, cracks were observed in the deep-water piers. These piers

developed cracks from the corner of the girder base plate and were propagated for great lengths.

The scope of this case study is to highlight the application of a newer generation strut-and-tiemodel, which was not in practice at the time of the original design. Thus these piers were not

designed with adequate reinforcement and therefore remedial post tensioning was required.

Depth to span ratios vary from 1 to 2 and girders are transferring loads very close to the

support edge, making these hammerheads ideal candidates for STM application. There could be

numerous reasons for the cracks to develop. Shrinkage, stress concentration or some erection

condition may be a few of them.

During STM analysis, presence of cracks was not considered but the existence of the crack will

redistribute the stress flow. The choice of load path is limited by the deformation capacity of the

beam and a situation may arise when a structure is unable to undergo the force distribution to reach

the assumed load path due to presence of cracks. One of the three hammerhead pier caps modeledin STM for this study is shown in Fig. 6.

1. Pier Cap 1- Length 28-ft, width 4-ft, depth at the end 3-ft 6-in and at the pier face 9-ft, 4-loads 250 kip each are placed on the top of the cap. The first load is 2-ft from the left endand then the rest are at 8-ft intervals. So the last load is 2-ft from the right end.


7/15


2. Pier Cap 2- Length 28-ft, width 5-ft, depth at the end 4-ft 6-in and at the pier face 14-ft,4-loads 290 kip each are placed on the top of the cap. The first load is 2-ft from the left

end and then the rest are at 8-ft intervals. So the last load is 2-ft from the right end.

3. Pier Cap 3- Length 28-ft, width 6-ft, depth at the end 6-ft and at the pier face 28-ft, 4-loads 550 kip each are placed on the top of the cap. The first load is 2-ft from the left end

and then the rest are at 8-ft intervals. So the last load is 2-ft from the right end.

As per this case study 7.5 sq-in reinforcements at the top tie level provided acceptable strength forall three hammerheads.

There could be numerous reasons for the cracks to develop. Shrinkage, stress concentration or

some erection condition may be a few of them. During STM analysis, presence of cracks was not

considered but the existence of the crack will redistribute the stress flow. The choice of load pathis limited by the deformation capacity of the beam and a situation may arise when a structure is

unable to undergo the force distribution to reach the assumed load path due to presence of crack.

In connection with the crack, the common retrofit is post-tensioning. In the strut-and-tie method,

the external post-tensioning can be efficiently modeled as external load. All force acting on the

anchorage zone shall be considered in the selection of a strut-and-tie model which should follow apath from the anchorages to the end of the anchorage zone.

Case Study 5 Curved Concrete Box Girder Bridge: Concrete box girder bridges can be

adapted to curved alignments where the box cross section is rigid torsionally. The effect of torsion

in a box girder bridge is influenced by the radius of curvature, span lengths, out-to-out width of box

structure, depth of structure, and thickness of deck, soffit, and exterior girder webs.

There are no simple rules of thumb to determine whether torsion is a significant factor to be

considered in the design of any particular curved structure.

The softened truss model theory applied to reinforced concrete or prestressed concrete

multiple cell box was developed by Fu, et al (Fu and Yang, 1996; Fu and Tang, 2001). By using

this model, the concrete torsional problem is solved by combining equilibrium and compatibilityconditions and constitutive laws of material, which is more sophisticated than the pure equilibrium

STM model.

A schematic sketch of a multi-cell box section under torsion is shown in Fig. 7(a) while

simplified forces subjected to shear, torsion, and bending are shown in Fig. 7(b). A reinforced or

prestressed concrete element, as shown in Fig. 7(c), is reinforced orthogonally with longitudinal

and transverse steel reinforcements. After cracking, the concrete is separated by diagonal cracks

into a series of concrete struts, as shown in Fig. 7(d). This example shows that the STM can also

solve the problem of girder under torsion.

Comparison of AASHTO and ACI 318 Provisions - Table 1 shows examples of stress limits and

strength reduction factors defined in the ACI Code and AASHTO LRFD Bridge DesignSpecifications, respectively. As shown in the table, there are substantial differences in the rules

used in these provisions and guidelines because of uncertainties associated with defining the

characteristics of an idealized truss within a continuum of structural concrete.

SUMMARY

The five cases shown above demonstrate that whenever common practice is used for designing D-

Regions, the practice leads to deficiencies or inefficiencies in the design of these commonly


8/15


occurring and often critical parts of structures. Due to the inadequacies in common practice,

coupled with the unlimited variety of D-Region shapes and loading conditions, it is not surprising

that most structural problems occur in D-Regions.

Our findings of STM Models used in the above-mentioned cases studies are:

The STM formulation that requires the least volume of steel will be the solution that bestmodels the behavior of a concrete member

This approach holds great promise for DOTs and design offices which could develop orobtain standard STMs for certain commonly encountered situations

Standard reinforcement details based on an STM could be developed for commonsituations

The STM then could be reviewed and revised if any parameters change

The CAST program developed by Kuchma is a useful tool and proved the previousfindings when the projects were conducted.

REFERENCES

AASHTO,AASHTO LRFD Bridge Design Specifications, American Association of State Highway

and Transportation Officials, Washington, DC, 1994 (1st Ed.), 1998 (2nd Ed.), and 2004 (3rd Ed.)

ACI Committee 318,Building Code Requirements for Structural Concrete (ACI 318-02) and

Commentary (ACI 318R-02), American Concrete Institute, Farmington Hills, Michigan, 2002

Collins, M.P. & Mitchell, D., (1991) Prestressed Concrete Structures, Prentice-Hall, Englewood

Cliffs, N.J.

Fu, C. C. (1994) Study of Crane Beam Check by using the Strut-and-Tie Model, an internal

study report to the Maryland Port Administration, University of Maryland, College Park, MD

Fu, C. C. & Yang, D., (1996) Designs of Concrete Bridges with Multiple Box Cells Due toTorsion Using Softened Truss Model, ACI Structural Journal, Technical Paper, Vol. 94, No. 6,

Nov-Dec

Fu, C. C. & Tang, Y., (2001) Torsional Analysis for Prestressed Concrete Multiple Cell Box,

ASCE Journal of Engineering Mechanics, Vol. 127, No. 1, January

Hsu, T. T. C., (1993) Unified Theory of Reinforced Concrete, CRC Press, Inc., Florida

Kuchma, D., CAST (Computer-Aided Struct-and-Tie) program downloaded from web site

http://www.cee.uiuc.edu/kuchma/strutand_tie/ (Version 0.9.11, Last Update: January 26, 2004)

MacGregor, J.G. (1997) Reinforced Concrete Mechanics and Design, 3rd Edition 1997,

Prentice-Hall

Marti, P., "Basic Tools of Reinforced Concrete Beam Design,"ACI Journal, Proceedings, Vol. 82,

No. 1, January-February 1985, pp. 45-56

Schlaich, J., Schfer, K., and Jennewein, M., "Toward a Consistent Design of Structural Concrete,"

Journal of the Prestressed Concrete Institute, Vol. 32, No. 3, May-June 1987, pp. 74-150


9/15


Table 1 Comparison of ACI Code and AASHTO LRFD Bridge Design Specifications

AASHTO LRFD ACI 2002

Struts

=

= smallest angle between the strutunder review and the adjoining ties

= average tensile strain in the tiedirection

= specified concrete compressive

strengthNote: The stress limit that assumes a

minimum distributed reinforcement of

0.003 in each direction is provided.

for prismatic struts in uncracked

compression zones

for struts in tension members

struts may be bottle-shaped andcrack control reinforcement is included

struts may be bottle-shaped andcrack control reinforcement is not included

for all other cases

specified concrete compressive strengthNote: Crack control reinforcement

requirement is , where

= steel ratio of the i-th layer of

reinforcement crossing the strut under

review, and = angle between the axis ofthe strut and the bars.

Nodes

when nodes are bounded by strutsand/or bearing areas

when nodes anchor only one tie

when nodes anchor more than one

tie

when nodes are bounded by struts

and/or bearing areas

when nodes anchor only one tie

when nodes anchor more than one

tie

Resistance

Factorfor struts and nodes

for ties

for struts, ties, and nodes


10/15


Figure 1 B- & D-Regions for Various Types of Members


11/15


2a Elevation Drawing

2b STM Model

Figure 2 Case Study 1 - Abutment on Pile


12/15


3a Elevation Drawing

3b STM Model

Figure 3 Case Study 2 - Walled Pier


13/15


Figure 4 - Gantry Crane Beam at Maryland Port Authority Harbor

72"

42.87"60.04"

5 SPANS @ 72"

42.87" 60.04" 42.87" 60.04" 42.87"

180K180K 180K 180K 180K 180K 180K 180K

Figure 5 Schematic Sketch of the Gantry Crane Beam


14/15


(a) Schematic View (b) Front view of one of the deep girders

Figure 6 Case Study 4 - Hammerhead Pier Type 3 of Thomas Jefferson Bridge


15/15


(a) Schematic Sketch (b) Space Truss Analogy

(c) Shear element (d) Truss element

Figure 7 Case Study 5 - Strut-and-Tie Model for Girder under Torsion

maryland experience in using strut-and-tie model in infrastructures

Documents